About vessel design temperature 5

[adx]

Hi! There is a problem I am suffering with in fire exposure scenario:

A PSV to service product cut tank, the bubble point of bulk liquid @ MAWP is far higher than MAWT, how should I do ?

More details, the MAWT of product cut tank is 205C, but the PSV/RD action temperature @ MAWP (1barg) is 280C.

What is the optional solution ? TKS !

Best Regards
ADX

 

[CJKruger]

adx, I normally make the design temperature high enough to cover all relief cases, EXCEPT the fire case. Not sure if this phylosophy is backed up by wording in the code.

 

[SeanB]

It is not a normal practice to consider the fire case relieving temperature when setting design temperatures. When setting your design temperature you consider your worst case operating temperature or upset case and then add your margin above. You do not consider the fire case temperature.

 

[JoeWong88]

adx,

Is your storage tank designed to pressure vessel code ? With such a high operting temperature, i guess it is. Please advise.

What are scenario may potentially adding heat into your system and results bulk liquid temperature reach bubble point (280 degC) ?

Potential candidates for heat input scenarios would be exothermic reaction, internal heater (if any) or external fire (may be more...)

I would discuss exothermic reaction here as it is too complicated to discuss...

For internal heater, there is a possibilit of heater controller failured results continuous heating te bulk liquid and reach the bubbling point. One way is to raise the design temperature as high as 280 degC. Then what happen after the temperature reach 280degC ? Si the fluid having wide boiling temperature and bulk fluid temperature continue to rise ? With this, you need to find the worst possible boiling temperature of the component in order to set the design temperature. Let say you are storing heavy crude, the temperature would be > 500 degC. If we set the design temperature at >500 degC for a storage tank, it will be a VERY THICK WALL tank. Are we going to do that ? My experience is NO. Then how to tackle high temperature due to heater failure ? One of the way is to provide highly reliability (High SIL) temperature sensing and power cut-off system and conduct SIL study to eliminate this demand on this case.

For fire case, regardless what design temperature you set (i hope nobody will set the design temperature as the fire flame temperature), the tank will eventual fail due to external fire. This inline with CJKruger and SeanB's statement and recommendation in API STD 521, section 5.15.1.1. Additional measures as stated in section 5.15.4 should be considered.

If you are interested, read this for your knowledge "Should maximum recommended wall temperature (Tw) for carbon steel vessel used as design temperature ?"

 

[adx]

Thank you, CJKruger and SeanB and especially JoeWong88 !!

Maybe I need describe more detailed about what I faced with

the design temperature of vessel is 205C, and normal operating temperautre is 160C, which vessel is product cut tank from full vacuum collumn due to heavy boiling temperature.

This vessel has PSV to protect from fire exposure. I run my company's tool to do the static and dynamic calculation, which just use API521 and some other mathematics tools.

So I get the results that:

1) this temperature of vessel would reach MAWT after fire exposure 30 min @ 0.5Barg (it means that PSV or RD would not act due to MAWP=1barg)

2) This temperature of vessel would reach 280C when PSV/RD is broken@1barg to protect this vessel overpressure. It spends 90 mins of fire exposure.

3) So in this simulation, I find that the vessel structure maybe failure before PSV/RD providing the protection....

4) In my opinion, add 50mm insulation for fire protection, everything is done. But it is disagreed, and we are trying to find something to let the calculation result reasonable.

How about your opinion and solution ?

Thank you so much !

Best Regards
ADX

 

[JoeWong88]

Providing external insulation is one of the way to minimize the impact of the fire attack. This is pretty inline with API Std 521 recommendation.

However, there are several problems associated with insulation :

i) The insulation external protection layer shall be able to resist high temperature ( > 900 degC per API) and sustain fire for 2 hours. Please check with you safety engineer and mechanical engineer for this compliance.

ii) Corrosion under insulation is a complicated and tricky stuff. It is hard to avoid and you need corrosion protection layer. The worst thing is operator may not aware of the occurrence of corrosion by visual inspection. A sophisticated assessment and inspection method is required. In the event, corrosion occur, there is potential of the corrosion material accumulated and damage the insulation layer without knowing. It may defect the purpose of insulation.

iii) The protection layer and insulation is subject to the risk of external mechanical damage during maintenance, inspection, site modification works, etc. Once the layer is damage, it is very difficult to maintain similar level of protection as per original requirement such as fire resistance for 2 hours at > 900 degC. Another tricky situation.

I do encourage you to reconsider the use of insulation.

I also suggest to look at the fixed foam-water spraying system and with high water density, fixed foam pouring system for bundled area, dedicated monitor for these tanks, design proper drainage system, enhanced fire detection system, etc.

 

[devaxrayz]

All,

I'm sorry but I’m a little confuse here and maybe this will be a little bit out of discussion. First question came over of my head IS DESIGN TEMPERATURE ALWAYS THE SAME WITH MAWT?

You might already hear the similar question with term of pressure and the answer is no (because it depends on what is the wall thickness you have from the manufacturer).

FIRST, Please correct me if I'm wrong, but my understanding of MAWT is the temperature where the vessel material will fail by its internal (design) pressure because it strength is reduced due to elevated temperature. And base on my understanding above, your MAWT will be depends on what is the thickness of your vessel out from the manufacturer. It may not the same (should be same or higher) compare with your design temperature since the wall thickness choose may be thicker due to material available in the market or shop or some conservative calculation roundup.

SECONDLY, this is only my learning thought , I assume that your vessel material is steel (carbon or stainless). The vessel thickness is calculate base on formula
t=pressure x diameter/(4 x allowable stress x joint eff.) plus corrosion allowance. One of the role of temperature in vessel design is that the allowable stress is reduced when the temperature is higher. However in design book for temp. -20F to 650 F (-29C to 343 C) the allowable stress is the same. It means whether you put 205 C or 280 C as your design pressure, it will give you the same thickness then the same strength. So your design pressure is 205 C but your MAWT is up to 343 C. AM I THINKING RIGHT?? or might it be that the manufacturer use higher allowable stress that not quote on vessel design textbook for their design then resulting in thinner wall thickness?

I have read one of the literature that said since the allowable stress is the same from -20F to 650F for steel, its better to put your design temperature up to 650 F since it would not give you more additional cost except additional requirement for gasket and cost due to possibility to increase your flange rating.

NEED ANY ADVICE

regards,


-rayz-

 

[carol2005]

I've learnt a lot by reading the discussions. I ran into a similar question. My plant owned a grandfathered pressure vessel designed for 625 PSIG and full vacuum, design temperature was 500 F. It is certainly adequate for normal operation condition of 450 PSIG @ 20 C ; However, the vessel is used as a purifier and the packing in the vessel requires regenaration with hot N2 purging for 2 hours every a couple of days. The temperature for the regenaration process could exceed the design temperature 500F. This is a debate wether or not it is a non-compliance with code requirement since we operate above the design temperature during regenaration process. I think it is fine because the regenaration pressure is very low, only 20 PSIG. But I might be wrong. Please advise.

 

[moltenmetal]

adx: it's the same with gas-filled vessels and vessels filled with very high boiling liquids when they're fitted with re-sealing relief devices as opposed to rupture disks. What you need to have at the end of the day is a safe fire case, meaning that the fire isn't likely to cause the vessel to rupture. There are several ways to accomplish it in your case:

- increase the MAWT to permit the fluid to boil safely and relieve by pressure during the fire
- supplement pressure relief with thermal relief
- eliminate the vessel from the fire case by changing its location etc.

All fire-clad insulation does is reduce the heat input rate during the fire, limiting the relieving rate once the PSV opens. The reduced heat input rate may give you time, but unless you can be sure the fire will be out before this time is elapsed, you might be in trouble.

Depending on what your tank is made out of, 205 C and 280 C may not actually be significantly different in terms of the vessel design. Once the calcs are done, you may merely need to have the vessel re-rated for a higher temperature at the same MAWP, or perhaps the calc will require you to drop your PSV setting from 1 bar(g) to 0.9 bar(g). Or perhaps nothing will be required: some owners permit the use of higher stress values in design for a fire relief event, under the undstanding that the unit may not be put back into service after the fire: ie. the vessel is designed such that fire may damage it by yielding it, but not (likely) causing it to fail. This approach is used sometimes with gas-filled vessels.

 

[CJKruger]

devaxrayz, maybe a slightly different subject. But I absolutely hate the practice of using a design temperature of at least 650 F.

People say the stress value doesn't change and it may save you money in future, but this is misleading. The fact is that it causes a lot of confusing and adds cost, because:
- Flange ratings drop long before you get to 650 F.
- Many instruments and miscellaneous attachments (like bushings) on the equipment cannot be rated that high. So either you have to add huge expense, or have two design temperatures (very confusing).

 

[JoeWong88]

rayz,

I'm sorry but I’m a little confuse here and maybe this will be a little bit out of discussion. First question came over of my head IS DESIGN TEMPERATURE ALWAYS THE SAME WITH MAWT?

Design temperature can be same as MAWT but shall not exceed.

A) MAWT
Here we must be a bit careful when we say MAWT.
MAWT for tank is subject to vessel/tank wall thickness... MAWT for a flange, instrument, attached to it may be much lower than the MAWT of tank.
We shall always refer MAWT for SYSTEM instead of a equipment / instrument.

Why design temperature ? Why MAWT ?
This main due the MAWT for equipment, instrument, piping, etc are different. However, ALL MAWT shall equal or more than the design temperature. Design temperature is what process demand, MAWT is what the equipment / device can take.


B) Coincident design pressure and pressure
Another concept we shall keep in mind is COINCIDENT design pressure & temperature.

In many cases, the fluid is operate at high pressure but low temperature (e.g. 16 barg with 40 degC). Process engineer may specify design condition of 18 barg @ 70 degC. However, the equipment shall be steam-out with LP steam (saturated steam 3.5 barg @ 150 degC). Thus the design condition shall be

Condition 1 : 18 barg @ 70 degC
Condition 2 : 2 barg @ 180 degC

The equipment, instrument, flange, piping, fitting, etc within the system shall be good for BOTH conditions BUT NOT necessary WORST of BOTH.


Carol2005,
Your system is exactly fit into the Coincident design condition case. (high P @ low T and low P @ high T)

BUT you shall always keep in that after regenaration, the vessel is still HOT. You SHALL NOT pressurize the system again before the vessel is cold below 500 F. Your system may experience coincident high P & T.


As mentioned earlier, for fire case, regardless what design temperature you set, the tank will eventual fail due to external fire. Additional protective measures shall be considered. Read more in "Protective Measures against FIRE other than Pressure Relief Device (PRD)"


Hope this make sense.

 

[devaxrayz]

CJKrueger,
Your saying about flange and instrumentation is absolutely right, and I agree that we can not take the 650 F design temperature thing easily.

Joe Wong,

thanks for your explanation, it is confirmed my understanding about MAWT. So back to adx question,

So I get the results that:

1) this temperature of vessel would reach MAWT after fire exposure 30 min @ 0.5Barg (it means that PSV or RD would not act due to MAWP=1barg)

2) This temperature of vessel would reach 280C when PSV/RD is broken@1barg to protect this vessel overpressure. It spends 90 mins of fire exposure.

3) So in this simulation, I find that the vessel structure maybe failure before PSV/RD providing the protection....

4) In my opinion, add 50mm insulation for fire protection, everything is done. But it is disagreed, and we are trying to find something to let the calculation result reasonable.

it can not be say as statement (1) since his MAWT maybe higher than his design temperature. It need calc first before jumping to conclusion (3) and recommendation (4).

I was agree with Joe opinion that whatever the design temperature, you will most likely fail your vessel in fire case. Maybe at inside the vessel your temperature still acceptable because heat is transferred to liquid inside and boil it, but at outside, the wall that interfacing the flame will easily exceed the design temperature. That is why a fire water system is put in place, one of the purpose it to keep your vessel cool enough during fire case.


-rayz-

 

[carol2005]

Thank you, Joe Wong, for the advice. After the packing in the vessel is hot purged for 2 hours, the operation procedure requests to put the hot N2 on cooling, slowly reduce the N2 flow temperature to norminal 20 C.

The debate was if operation was to run above the vessel design temperature during regen process, is there a legislative consequence? Is it allowed by legislation governing codes? I don't know enough about the pressure vessel codes to defog that concern.

 

[JoeWong88]

rayz,

it can not be say as statement (1) since his MAWT maybe higher than his design temperature. It need calc first before jumping to conclusion (3) and recommendation (4).

Well...i don't exactly know what adx has done to estimate the heating time & temperature. I just have to assume what adx has done is fit, good, make sense, reliable, compiance with code, etc. Point (1) to (3) are the results and observation from studies. Based on the results and observation, adx has taken the external insulation approach. So it make sense and in compliance with API.

adx may have taken the MAWT of the system (lowest out of all MAWt of vessel wall, instrument,...)...

carol2005,

The debate was if operation was to run above the vessel design temperature during regen process, is there a legislative consequence? Is it allowed by legislation governing codes? I don't know enough about the pressure vessel codes to defog that concern.

I am not specialist in vessel design and in contact with authority. So can not advise nor confirm anything... But my experience is the design condition should be declared clearly and it has been tested according to design code requirement before it is in operation. Persoanlly i would guess this has been done.

I also hope that your system has built-in interlock to avoid mal-operation, otherwise a proper checklist and procedure (provide multiple safety gates)shall be in place to ensure NO mal-operation.

Hope you enjoy reading.